Find a combination of parameters that best re-creates the data. Draw from the posterio and observe which simulation best matches
Take the parameters that provided the best result above, and simulate using only those parameters to see how much “natural” stochasticity there is in the timing of the outbreak.